Introduction¶
This report presents a geospatial analysis of MRT accessibility across Singapore's planning subzones. Using demographic, economic, and transport datasets, the study aims to identify underserved areas and propose improvements for MRT infrastructure planning.
Data Sources¶
- Passenger Volume by Train Stations (April 2025)
- Resident Households by Planning Area of Residence and Monthly Household Income from Work (Census Of Population 2020)
- Employed Residents Aged 15 Years and Over by Planning Area of Workplace, Usual Mode of Transport to Work and Planning Region of Residence (Census of Population 2020)
- Resident Households by Planning Area of Residence and Tenancy (Census Of Population 2020)
- Resident Population by Planning Area/Subzone of Residence, Age Group and Floor Area of Residence (Census of Population 2020)
- Employed Residents Aged 15 Years and Over by Planning Area of Workplace and Usual Mode of Transport to Work (Census of Population 2020)
- Master Plan 2019 Planning Area Boundary (No Sea) (GEOJSON)
- Master Plan 2019 Subzone Boundary (No Sea) (GEOJSON)
Methodology¶
- Cleaned and merged spatial and tabular datasets using GeoPandas and Pandas
- Created plots and maps to visualize Singapore's population, population density, average monthly income, MRT lines and other socioeconomic metrics
- Computed distance from subzones to nearest MRT station using geodesic distance
- Identified underserved subzones using multiple factors (population density, working population, MRT user ratio, MRT distance)
- Scaled features for clustering (population density, working population, MRT user ratio, MRT distance)
- Performed KMeans clustering to group subzones based on transport and demographic characteristics
- Visualized results using Plotly and geopandas for interpretability
- Identified subzones that require improvement in MRT accessibility and prescribed possible solutions for each
Average Monthly Household Income by Planning Area¶
This map and bar chart illustrates the distribution of average household incomes across Singapore’s planning areas. High-Income Areas:
- From the bar chart, it is visually evident that the planning areas Tanglin, River Valley, Bukit Timah, and Downtown Core stand out dramatically, with average monthly household incomes exceeding $13,800. These are largely affluent central districts, often with a concentration of private housing or expatriate communities.
Middle-Income Areas:
- Areas like Sengkang, Punggol, Choa Chu Kang, and Clementi fall in the $9,000–$11,000 range, indicating established or newer HDB towns with a mix of public and private housing.
Lower-Income Areas:
- Outram, Ang Mo Kio, Toa Payoh, and Bukit Merah rank lowest in average income, with Outram below $7,000/month. These often contain mature estates with aging populations and smaller household sizes.
Implications for MRT Planning:
Affluent areas often have better access to private transport, so MRT planning may not need to prioritize them.
Middle-income and underserved areas (like Choa Chu Kang or Sembawang) may benefit more from enhanced MRT connectivity, especially if income does not correlate with access.
Lower-income zones such as Outram and Ang Mo Kio may be more reliant on public transport, suggesting strong justification for improving MRT accessibility or reducing congestion in these areas.
Population by Planning Areas and Subzones¶
These maps and charts helps in identifying macro-level population clusters, which is crucial for determining where MRT services should be concentrated, expanded, or optimized.
High-population planning areas such as Bedok, Jurong West, Tampines, Woodlands and Sengkang stand out clearly, indicating areas with greater residential demand and possibly higher MRT ridership potential.
Some planning areas have very low population like Downtown Core, Chnagi, Singapore River, Tuas, Sungei Kedut, etc are because these areas are business districts/ industrial areas with low residential population.
Some planning areas are reserved / heavily forested with few population, like Mandai, Central Water Catchment and Western Water Catchment.
As this data is extracted from the 2020 Census, Tengah has not fully developed yet.
Population Density by Planning Area¶
These maps visualize population density (residents per km²) at the planning area and subzone level. Population density at subzones level allow for a more nuanced spatial understanding of where residents are most densely concentrated.
MRT Lines and Stations¶
This map visualizes all current MRT lines and stations in Singapore, including stations under construction such as those on the Cross Island Line (CRL) and Jurong Region Line (JRL). It provides a comprehensive overview of both existing and future rail infrastructure.
Existing MRT lines (e.g., East-West, North-South, Downtown) are shown alongside planned expansions.
Future stations are included based on available data.
This map is used to calculate the distance of a Subzone's centroid to the nearest MRT distance to approximate the accessbility of MRT services.
This helps reveal network gaps, where densely populated or highly employed areas may still lack direct MRT access even after upcoming expansions.
Employment Density by Planning Area¶
Employment density refers to the number of jobs or employed persons per unit area (e.g., per square kilometre) in each planning area. It highlights where economic activity is concentrated and where large numbers of people work daily.
Central Business District (CBD) areas such as Downtown Core, Orchard and Outram show the highest employment densities due to a high concentration of commercial buildings, offices, and government institutions.
Although other planning areas have relatively lower employment density, this is maybe due to their larger size, the employment population map can also be used to gauge other planning area's economic activity.
Employment density is a key indicator for transport demand during peak hours. Areas with high employment density often face congestion issues, particularly at MRT stations during the morning and evening rush.
It helps evaluate if MRT coverage is sufficient in high-employment areas and whether stations are well-placed to support commuting patterns.
MRT Usage Rate Among Employed Residents by Planning Area¶
This metric reflects the proportion of employed residents in each planning area who rely on the MRT as part of their daily commute to work. It is calculated as:
MRT Usage Rate = (Number of people using MRT or MRT+combination to commute) / (Total number of employed persons)
It includes:
Those who commute only via MRT/LRT
Those who commute via MRT/LRT combined with public bus or other modes
Based on the calculated MRT usage rate — which measures the proportion of employed residents in each planning area who use the MRT (either solely or in combination with other modes) to commute — the three lowest-ranked areas are:
Pasir Ris – 21.2%
Seletar – 25.2%
Bukit Panjang – 28.9%
Possible Reasons:
- Pasir Ris (21.2%)
Despite being served by the East-West Line, it is located at the easternmost tip of Singapore, making it geographically far from key employment hubs like the CBD or Jurong.
Commute times may be long, and residents might rely on private transport or bus services instead.
The Cross Island Line, currently under construction, is expected to improve accessibility.
- Seletar (25.2%)
A low-density residential and industrial area with no direct MRT access as of now.
Residents typically depend on private cars, buses, or shuttle services to nearby stations (e.g., Yio Chu Kang or Sengkang).
- Bukit Panjang (28.9%)
Though served by the Downtown Line and Bukit Panjang LRT, the overall connectivity is still limited compared to other mature towns.
Transfer friction from LRT to MRT and long travel times to city centers may reduce attractiveness.
MRT usage may rise with ongoing improvements in network integration and feeder connectivity.
A low MRT usage rate in an area with high employment density suggests a gap in accessibility, signaling a need for improved MRT coverage, shuttle services, or better last-mile options.
High MRT usage rates validate the success of existing MRT coverage and planning, but also point to potential capacity stress during peak hours.
Comparing usage rates with tap-in/tap-out ratios helps to triangulate commuting behavior, indicating source-destination patterns and transit bottlenecks.
[NbConvertApp] Converting notebook Singapore_Overview.ipynb to html C:\Users\Chia Hon Lam\AppData\Local\Programs\Python\Python310\lib\site-packages\zmq\_future.py:718: RuntimeWarning: Proactor event loop does not implement add_reader family of methods required for zmq. Registering an additional selector thread for add_reader support via tornado. Use `asyncio.set_event_loop_policy(WindowsSelectorEventLoopPolicy())` to avoid this warning. self._get_loop() [NbConvertApp] Writing 17788898 bytes to Singapore_Overview.html